AMD's Radeon HD 5870: Bringing About the Next Generation Of GPUs
by Ryan Smith on September 23, 2009 9:00 AM EST- Posted in
- GPUs
A Quick Refresher on the RV770
As Cypress is a direct evolution of the RV770 design, before we talk about what’s new with Cypress we are going to go over a quick rehash of RV770’s internal workings. As it’s necessary to understand how RV770 was built to understand what Cypress changes, if you’re completely unfamiliar with RV770, please take a look at our expanded discussion of RV770 from last year. For the rest of you, let’s get started.
At the center of the RV770 is the Stream Processing Unit (SPU), a single arithmetic logic unit. The RV770 has 800 of these, and they are packaged together in groups of 5 and are what we call a Streaming Processor (SP). A SP contains a register file, a branch predictor, and the aforementioned 5 SPUs, with the 5th SPU being a more complex unit capable of transcendental functions along with the base functions of an ALU. The SP is the smallest unit that can do individual work; every SPU in an SP must execute the same instruction.
For every 16 SPs, AMD groups them together with texture units, L1 cache, shared memory, and controlling logic. This combined block is what AMD calls a SIMD, and RV770 has 10 of them. These 10 SIMDs form the core computational power of the RV770, and in the chip work with various specialized units such as ROPs, rasterizers, L2 cache, and tesselators to form a complete chip.
To utilize the computational power of the hardware, instruction threads are issued to the SPs. These threads are grouped into wavefronts, where there are 64 threads per wavefront. To maximize the utilization of the GPU, threads need to be organized so that they can feed all 5 SPUs in a SP an instruction every clock cycle. Doing this requires extracting instruction level parallelism (ILP) out of programs being passed to the GPU, which is difficult task of AMD’s compiler.
If SPUs go unused, then the performance of the chip suffers due to underutilization. This design gives AMD a great deal of theoretical computational power, but it is always a challenge to fully exploit it.
Facebook
Twitter
RSS
327 Comments
View All Comments
T2k - Thursday, September 24, 2009 - link
Just whatF are you talking about, seriously? Is this some kind of new urban BS again?
Look at other sites, they DID test it with ET:QW:
http://www.techpowerup.com/reviews/ATI/Radeon_HD_5...">http://www.techpowerup.com/reviews/ATI/Radeon_HD_5...
http://www.xbitlabs.com/articles/video/display/rad...">http://www.xbitlabs.com/articles/video/display/rad...
and so on. BTW L4D is a passing fart in the wind while ET:QW is still going strong, stronger than UT3 (unfortunately because UT3 looks 10x better and the old ONS mode was awesome but Epic fucked it in UT3)
Read other reviews before you post:
http://forums.anandtech.com/messageview.aspx?catid...">http://forums.anandtech.com/messageview...amp;thre...
SiliconDoc - Wednesday, September 30, 2009 - link
I see by your review links the 5870 just doesn't do well in ET:QW, everything else is closer to it and it gets beat worse, at 2560 high aa&af GTX295 slams it by 20%, so, of course, it was left out, in order to "pump up the percieved red number".-
It's "not a good appearance".
You'll just have to live with it for now like the rest of us.
number58 - Wednesday, September 23, 2009 - link
Is this card any closer to saturating the pci-express 2.0 x16 slot? That was one of the big arguments in the P55 vs X58 debate. Would there be any loss of performance using these in crossfire on P55 compared to X58?Otherwise, great article. I think I'll be hanging on to my 4890 for a while though.
Kaleid - Wednesday, September 23, 2009 - link
Look here for an answer:http://www.techpowerup.com/reviews/AMD/HD_5870_PCI...">http://www.techpowerup.com/reviews/AMD/HD_5870_PCI...
number58 - Wednesday, September 23, 2009 - link
Thanks. That just about settles it. My next system will be Lynnfield based.Zeratul - Wednesday, September 23, 2009 - link
Please add support for Stereoscopic 3D, ATI. I'm not going back to play games in 2D, though I don't like nvidia's monopoly in that.Arbie - Wednesday, September 23, 2009 - link
Not just a review but an excellent technical discussion. It must have been a lot of work. Thanks for all of that, and do please follow up where you can on people's requests for more info on this very important card.Kudos to ATI/AMD for such an achievement. It looks like Nvidia is in big trouble.
Arbie
SiliconDoc - Wednesday, September 23, 2009 - link
rofl - Nvidia is in big trouble ? LOLHave you looked at the charts on wiki that predicted what this review shows ? Maybe you should take a gander at the Nvidia chart, now, and cry.
--
It is a nice review, although biased red here as usual.
--
I'd like to mention as he fretted about heat because of the 1/2 vent on the back of the card, I didn't notice overt despairing mention of the TWO rectangular exhaust ports on the fan end BLOWING HEAT INTO THE CASE.... which of course is deserved. Nice how that was held back as if overlooked. (Boy what trouble for Nvidia! hahha)
Next we'll be told the red rooster tester put his hand next to the two internal exhaust ports, and "the air felt rather cool" so "rest assured not much heat is being pushed into your case", and "this doesn't matter".
See, the trouble is already in the article, the trouble it took to spin it just so for Ati... lol... it's hilatrious!
ClownPuncher - Wednesday, September 23, 2009 - link
Bizzarro SnakeOil, is that you?SiliconDoc - Wednesday, September 23, 2009 - link
No, this is me, the same me I've always been.Would you like to comment on the two internal exhaust ports of the 5870 that put sweltering heat into your case ?
I guess you tried to ignore that entirely.
You can always insult me again and avoid commenting on the heating exhaust ports ramping up your case temps on the 5870...
That of course would be "the right thing to do".
Perhaps call me crazy, and avoid the topic, right, you're good at that, huh.